Pilot error brought down the Armies Watchkeeper

[NutLoose]

The drone was doing its thing up until human intervention scuppered it

https://www.theregister.co.uk/2019/08/13/watchkeeper_drone_crash_wk050_aberporth/

When it landed at Aberporth, WK050 "landed long", reported the BBC. This means that instead of touching down at the correct point towards the start of the runway, giving it a nice long distance in which to harmlessly roll to a stop, the drone was further along than it ought to have been – risking it overrunning off the far end, damaging the aircraft.Thus, the onboard computer followed its programming and "auto-aborted as it approached the end of the runway". The drone throttled up to full power and took off again, ready to fly itself around in a circuit and have another go at landing. Such things are a fact of life in aviation, whether humans or computers are trying to land.However, WK050's human operators seemingly became confused at this point – and cut the throttle. WK050 "glided over the road" at the end of the runway and "crashed into a tree" around 900 metres beyond the end of the runway."Had no action been taken by the crew the AV (aerial vehicle) would have completed its automatic go-around, from which it could have been commanded to conduct a further approach," the report said.

[Just This Once...]

The ongoing experiment with non-aircrew at the (virtual) controls of a fixed-wing aircraft would have made more sense if the Army had purchased a bunch of otherwise stripped-out, empty and suitably ballasted Hermes aircraft to play with & crash for the last 10 years or so. It would have saved quite a bit of money whilst providing similar operational effect.

[gijoe]

The ongoing experiment that sees RAF aircrew thinking they know all would make more sense if the 100+ year experiment was ended and the equipment given to the military branches of HM Forces. Have lost count how many times div, bde, pl, tp, section, range of mortar, etc has had to be explained to know-it-all-knobbers.

[betty swallox]

Dear chap. Are you suggesting that RAF crews were at the controls, or am I picking you up incorrectly?

[Audax]

Quite a rant there gijoe. When you’ve put your toys back in your cot (assuming you’re allowed toys), could you translate please so normal humans can understand?
​​​​​

[switch_on_lofty]

Have lost count how many times div, bde, pl, tp, section, range of mortar, etc has had to be explained to know-it-all-knobbers.

Probably wouldn't need to explain that cutting the engine at the start of an auto go around on a single engine aircraft wasn't a good idea though. Each to their own!
​​​​​

[WB627]

https://www.bbc.co.uk/news/uk-wales-49158509

https://ichef.bbci.co.uk/news/624/cpsprodpb/14791/production/_102075838_drone.jpg
This picture reminded me of flying "RPV'S" with my son, we never needed anything bigger than a bin bag to remove the wreckage and it went in the boot of my car, not on a flat bad

[CAEBr]

The ongoing experiment that sees RAF aircrew thinking they know all would make more sense if the 100+ year experiment was ended and the equipment given to the military branches of HM Forces. Have lost count how many times div, bde, pl, tp, section, range of mortar, etc has had to be explained to know-it-all-knobbers.
The ever decreasing number of watchkeepers are all operated by the British Army model flying club. The only involvement that the RAF have is when they are called upon to participate in the resultant Service Inquiry

[Wrathmonk]

Nice fishing exercise gijoe - a few bites. A few "whooshes" as well....!

Edited to add - we'll take the flying things, in exchange you can have the RAF Regiment and the RAFP :E

[higthepig]

The ongoing experiment with non-aircrew at the (virtual) controls of a fixed-wing aircraft would have made more sense if the Army had purchased a bunch of otherwise stripped-out, empty and suitably ballasted Hermes aircraft to play with & crash for the last 10 years or so. It would have saved quite a bit of money whilst providing similar operational effect.
Agree totally, because 2 winged demi god's have never smashed anything in, have they....

[Arthur1815]

Another lucid brown rant. I lost count of the number of times Rupert the donkey walloper failed to understand that the aircraft did not have to be parked on his front lawn for it to be effective at div, bde, coy ..........etc etc

[Bob Viking]

Please, God, tell me that was a serious post. I’ll be genuinely gutted if I find out it was a phishing attempt.

It made my day. Honestly.

BV

[Tankertrashnav]

Edited to add - we'll take the flying things, in exchange you can have the RAF Regiment and the RAFP https://www.pprune.org/images/smilies/evil.gif

Oi!!! No you dont!

TTN (ex Rock)

[AF03-111]

in all seriousness, this does raise questions about the quality and depth of the training being provided, and the airmanship (or lack of) being applied across the programme, The earlier WK crash as a result of deliberately flying the air vehicle into known icing conditions told me everything I needed to know about the lack of serious RPAS / UAV experience involved.

[Sideshow Bob]

"Pilot error brought down the Armies Watchkeeper" Just how many Armies do we have? Seriously though GIJoe does have a point, can you say with all seriousness that a RAF pilot has never fudged a go-around? I can think of serveral including a Tristar tail strike following a botched landing and go-around at Kandahar.

[Less Hair]

Shouldn't the operators have known what mode the drone is in and what it will do next? Sounds like some operator's training issue to me.

[higthepig]

in all seriousness, this does raise questions about the quality and depth of the training being provided, and the airmanship (or lack of) being applied across the programme, The earlier WK crash as a result of deliberately flying the air vehicle into known icing conditions told me everything I needed to know about the lack of serious RPAS / UAV experience involved.
It wasn't the Army flying that, it was Thales

[tucumseh]

It wasn't the Army flying that, it was Thales

Who issued the RTS, which is predicated on the correct training having taken place?

[higthepig]

Who issued the RTS, which is predicated on the correct training having taken place?
It was operating outside of the RTS at the time, it was Thales carrying out test flying, nothing to do with the Army.

[Chugalug2]

Never mind the Army, htp. Are you saying that Thales is operating under its own RTS, or under no authority at all?

[PPRuNeUser0211]

Never mind the Army, htp. Are you saying that Thales is operating under its own RTS, or under no authority at all?
Being a CFAOS organisation for watchkeeper I'd suspect, Thales would not be bound by the in-service RTS but would have their own (likely similar) document that may be more or less restrictive in various areas. If they were bound by the RTS it'd be pretty hard to do developmental test flying which, almost by definition, is doing stuff that is novel and untested and, therefore, not in the RTS...

[AF03-111]

If memory serves, Thales were flying the one that crashed in icing conditions under a military flight test permit which had been issued earlier that year. Whilst the the SI was (rightly) critical of the choices made such as flying the sortie with a rather expensive and limited-availability radar on board, the officers singularly missed the obvious point i.e. you don't deliberately fly an unmanned system into icing conditions, even more so in a system designed such that the Remote Pilots have no means of 'direct' control.

[Chugalug2]

Being a CFAOS organisation for watchkeeper I'd suspect, Thales would not be bound by the in-service RTS but would have their own (likely similar) document that may be more or less restrictive in various areas. If they were bound by the RTS it'd be pretty hard to do developmental test flying which, almost by definition, is doing stuff that is novel and untested and, therefore, not in the RTS...

Thanks for that, pba. Could you please say what CFAOS is, and what or who it is answerable to if not the MOD? AFAIK there are only two relevant Regulatory Aviation Authorities, the CAA/EASA and the MAA. Presumably Thales operates Watchkeeper testing under one or the other?

Of course testing means going outside of the Military RTS but presumably not recklessly, so was this accident the result of systematically testing the boundaries of the aircraft or simple carelessness? What was the role of the MOD Duty Holder in what happened, if any?

Thanks

[higthepig]

Thanks for that, pba. Could you please say what CFAOS is, and what or who it is answerable to if not the MOD? AFAIK there are only two relevant Regulatory Aviation Authorities, the CAA/EASA and the MAA. Presumably Thales operates Watchkeeper testing under one or the other?

Of course testing means going outside of the Military RTS but presumably not recklessly, so was this accident the result of systematically testing the boundaries of the aircraft or simple carelessness? What was the role of the MOD Duty Holder in what happened, if any?

Thanks
CFAOS (http://letmegooglethat.com/?q=CFAOS) and Watchkeeper SI link (https://aviation-safety.net › wikibase Web results Accident Thales Watchkeeper WK450 Mk 1 WK042, 03 Feb 2017)

[phil9560]

Its easily done.I had to climb a 20 foot tree to recover my drone last week.

[Chugalug2]

Thanks for the links htp. The SI one won't open for me but perhaps it is protected? The CFAOS one does, and contains this intriguing RA4051(1) in the sample Ops Manual:-

RA 4051(1): Flight Testing of Aircraft
Flight-testing and Airborne Checks (ACs) shall be conducted within the boundaries of the Release to Service (RTS).

[Easy Street]

Chug: ‘flight testing’ is what’s done after maintenance activity, colloquially known as an air test. Operating outside the RtoS is ‘test flying’.

[Chugalug2]

Chug: ‘flight testing’ is what’s done after maintenance activity, colloquially known as an air test. Operating outside the RtoS is ‘test flying’.

Ah, I thought it was too easy (sorry!). :O

Thanks for the explanation though ES.

[Bing]

the officers singularly missed the obvious point i.e. you don't deliberately fly an unmanned system into icing conditions, even more so in a system designed such that the Remote Pilots have no means of 'direct' control.

Then how do you propose they test the icing clearance? Which was one of the major selling points over the original Hermes 450.

[tucumseh]

Then how do you propose they test the icing clearance? Which was one of the major selling points over the original Hermes 450.

I think a typical Icing section in an RTS will provide a good indication of what factors are considered; some of which are universal (e.g. icing deemed to exist in certain temperature/visibility conditions), others aircraft-dependant (e.g. engine surge margins). Modelling will have given a good indication of what to expect, and one then gradually expands the opertating envelope to verify performance. But one tends not to (deliberately) get to the stage where the testing is destructive. Although things are a little blurred by today's use of the term 'drone', which is a quite different thing to people of a certain age!

[Bing]

I think a typical Icing section in an RTS will provide a good indication of what factors are considered; some of which are universal (e.g. icing deemed to exist in certain temperature/visibility conditions), others aircraft-dependant (e.g. engine surge margins). Modelling will have given a good indication of what to expect, and one then gradually expands the opertating envelope to verify performance. But one tends not to (deliberately) get to the stage where the testing is destructive. Although things are a little blurred by today's use of the term 'drone', which is a quite different thing to people of a certain age!

Oh the testing wasn't deliberately destructive. They were in the gradually exploring the envelope stage by my reading of the report, but didn't catch the deteriorating situation in time. I seem to recall they had most (all?) of the anti icing on though so probably not a pass for that data point.
Not that I don't think the whole programme hasn't been a massive cluster.

[tucumseh]

I've quickly scanned the recommendations in the two reports issued on the same day. Yep, seen them all before. Formal declarations made that the relevant regs had been met, when it seems much of the work hasn't started.

General Felton's rather pointed recommendations (WK043) amount to 'Now we've crashed 5 after the RTS was issued, let's think about issuing a legal RTS'. I wonder if he knows how many previous BoI/SIs have said this?

DG DSA (who certainly knows) excuses much due to complexity. The need for endorsed Aircrew and Operating Data Manuals isn't complex. It's bloody mandated. Incomplete understanding of how sub-systems integrate? I distinctly recall one person who later headed UAVs saying 'If it works on the bench, it'll work in the aircraft'. RTS signed with no integration, testing, trials, training, pubs, etc. Four years later, Tornado ZG710, 2 killed. At least the death count is now lower.

[AF03-111]

As Tucumseh said, quite a lot can be done with modelling but if it had been me making the decision, I would have insisted on LOTS of environmental chamber testing, including (if access could be gained) use of a Climactic Wind Tunnel. Beyond that, the process should have been incremental testing and "learning through experience" on training, test and operational flying, seeking to build a body of evidence over a period of time.
The problem with the "all weather" capability as marketed at the outset of the WK programme is that it wasn't something the manufacturer could actually deliver, and so they ended up trying to prove a capability that was in all likelihood going to result in the loss of the aircraft. The pressure to deliver may have been contractual i.e. the MOD insisting on it being tested to their satisfaction. It's bad enough being iced in a manned light aircraft, let alone a UAV that has positive 'pilot' control, and of course, WK is "point and click" so the operators could not intervene i.e. try to find a break in the clouds and get the hell out. Remember, the programme had already suffered an airframe loss due to water ingress / blockage in the pitot tube.
To anyone with a decent amount of UAV experience, putting the aircraft into these conditions was a pretty foolish thing to do and it suggests that those making the decision to do so either did not appreciate the risk, or did so but pressed on regardless.

[alfred_the_great]

As Tucumseh said, quite a lot can be done with modelling but if it had been me making the decision, I would have insisted on LOTS of environmental chamber testing, including (if access could be gained) use of a Climactic Wind Tunnel. Beyond that, the process should have been incremental testing and "learning through experience" on training, test and operational flying, seeking to build a body of evidence over a period of time.
The problem with the "all weather" capability as marketed at the outset of the WK programme is that it wasn't something the manufacturer could actually deliver, and so they ended up trying to prove a capability that was in all likelihood going to result in the loss of the aircraft. The pressure to deliver may have been contractual i.e. the MOD insisting on it being tested to their satisfaction. It's bad enough being iced in a manned light aircraft, let alone a UAV that has positive 'pilot' control, and of course, WK is "point and click" so the operators could not intervene i.e. try to find a break in the clouds and get the hell out. Remember, the programme had already suffered an airframe loss due to water ingress / blockage in the pitot tube.
To anyone with a decent amount of UAV experience, putting the aircraft into these conditions was a pretty foolish thing to do and it suggests that those making the decision to do so either did not appreciate the risk, or did so but pressed on regardless.

given that there was a (very) reduced risk to life, and that environmental testing may have cost more than a single airframe, why not fly it into icing conditions?

and it's boring for people to talk about risk without acknowledging that there is a concomitant opportunity. "Pressing on regardless" is not a bad thing - risks pay off (sometimes big) on a daily basis.

[Chugalug2]

"Pressing on regardless" is not a bad thing - risks pay off (sometimes big) on a daily basis.

Pressing on regardless by subverting the airworthiness regulations and illegally ordering their suborning is what did for UK Military Airworthiness Provision in the first place, and can be said to have lead to over 100 deaths in subsequent airworthiness related fatal air accidents.

This wasn't a battlefield situation requiring flouting of the RTS (such as sending a rescue mission clinging to the outside of an AAC Apache into a fire fight). This was a commercial company carrying out test flying of an MOD aircraft in UK airspace. I've no idea what the costs are vis a vis environmental chambers and total airframe loss, but I'm damn sure no such costing had been assessed and consequent action authorised anyway. The scandal of WK loss rates out of operational areas needs to be fully investigated. Pilot error doesn't begin to address the fundamental causes.

[tucumseh]

environmental testing may have cost more than a single airframe



I certainly remember when that was once true. As part of the RAF's 'savings at the expense of safety' policy of the late 80s/early 90s, a decision was made to cancel the ETC maintenance contracts. No amount of expert advice would persuade AMSO that they needed to be exercised regularly and, according to policy of the day, re-certified every year by the original manufacturer. Then along came EFA, wanting to use them... God knows how much it cost to resurrect them. Ferranti had a hangar-sized building full of them, all owned by MoD - but the plinths had rotted away and they leaked like No 10 DS. Cheaper to pay for the maintenance contract.